DHPC Bicelles Inside the Stratum

The effect of dipalmitoyl phosphatidylcholine (DPPC)/dihexanoyl phosphatidylcholine (DHPC) bicelles on the microstructure of pig stratum corneum (SC) ...
0 downloads 0 Views 2MB Size
5700

Langmuir 2008, 24, 5700-5706

Penetration and Growth of DPPC/DHPC Bicelles Inside the Stratum Corneum of the Skin L. Barbosa-Barros,*,† A. de la Maza,† J. Estelrich,‡ A. M. Linares,§ M. Feliz,§ P. Walther,| R. Pons,† and O. López† Departamento de Tecnologı´a de TensioactiVos, Instituto de InVestigaciones Quı´micas y Ambientales de Barcelona (I.I.Q.A.B.), Consejo Superior de InVestigaciones Cientı´ficas (C.S.I.C.), Calle Jordi Girona 18-26, 08034 Barcelona, Spain, Departamento de Fisicoquı´mica, Facultad de Farmacia, UniVersidad de Barcelona, AV. Joan XXIII s/n, 08028 Barcelona, Spain, SerVeis Cientificote`cnics, UniVersidad de Barcelona, Calle Baldiri Reixac 10-12, 08028 Barcelona, Spain, and Zentrale Einrichtung Elektronenmikroskopie, UniVersität Ulm, Albert-Einstein-Allee 11, D-89069 Ulm, Germany ReceiVed NoVember 29, 2007. ReVised Manuscript ReceiVed March 6, 2008 The effect of dipalmitoyl phosphatidylcholine (DPPC)/dihexanoyl phosphatidylcholine (DHPC) bicelles on the microstructure of pig stratum corneum (SC) in Vitro was evaluated. The physicochemical characterization of these nanoaggregates revealed small disks with diameters around 15 nm and a thickness of 5.4 nm. Upon dilution, the bicelles grow and transform into vesicles. Cryogenic scanning electron microscopy (cryo-SEM) images of the SC pieces treated with this system showed vesicles of about 200 nm and lamellar-like structures in the intercellular lipid areas. These vesicles probably resulted from the growth and molecular rearrangement of the DPPC/DHPC bicelles after penetrating the SC. The presence of lamellar-like structures is ascribed to the interaction of the lipids from bicelles with the SC lipids. The bicellar system used is suitable to penetrate the skin SC and to reinforce the intercellular lipid areas, constituting a promising tool for skin applications.

1. Introduction Bicelles are lipid nanostructures formed by long and short chain phospholipids dispersed in aqueous solution.1 These aggregates are normally used as membrane models for nuclear magnetic resonance (NMR) structural studies of membrane biomolecules.2 Depending on the composition and the long/ short chain phospholipid molar ratio (q), these structures may display a bilayered discoidal morphology alignable in magnetic fields at temperatures above the long chain phospholipid phase transition (Tm). The total lipid concentration of bicelles is also an important factor, since the dilution of these aggregates promotes morphological transitions.3 Considering the structural versatility, the lipid composition, and the bilayered assembly of the bicellar aggregates, these systems would be suitable for skin penetration; however, works on this subject are scarce. The stratum corneum (SC) of the skin constitutes the main target and the main barrier for transdermal drug delivery. The structure and dimensions of bicelles, with diameters in the range of 10-30 nm and a thickness of about 6 nm, are similar to those of liposomes and micelles. These two nanostructures are often used for skin research, although their application is in debate due to some aspects. Liposomes seem to be too large to penetrate into the narrow interlamellar spaces of the SC.4 Concerning the micelles, the presence of surfactant in their composition poses a problem due to the well-known irritating effect of these † Instituto de Investigaciones Quı´micas y Ambientales de Barcelona (I.I.Q.A.B.), Consejo Superior de Investigaciones Cientı´ficas (C.S.I.C.). ‡ Departamento de Fisicoquı´mica, Universidad de Barcelona. § Serveis Cientificote`cnics, Universidad de Barcelona. | Universitaet Ulm.

(1) Sanders, C. R.; Schwonek, J. P. Biochemistry 1992, 31, 8898–8905. (2) Whiles, J. A.; Deems, R.; Vold, R. R.; Dennis, E. A. Bioorg. Chem. 2002, 30, 431–442. (3) Bolze, J.; Fujisawa, T.; Nagao, T.; Norisada, K.; Saito, H.; Naito, A. Chem. Phys. Lett. 2000, 329, 215–220. (4) Benson, H. A. E. Curr. Drug DeliVery 2005, 2, 23–33.

solubilizing agents on the skin.5 In this context, the use of bicelles for skin treatment may be advantageous as compared to liposomes and micelles: the size of bicelles is small enough for passing through the SC lipid lamellae and their composition consists completely of lipids. In earlier works, we studied the structure and composition of the SC6–8 as well as the effect of different nanostructures and molecules, such as liposomes, lipid mixtures, surfactants, and organic solvents on the skin, particularly on the SC.9–11 Bicelles made up of dimyristoyl phosphatidylcholine (DMPC) and dihexanoyl phosphatidylcholine (DHPC) (molar ratio q ) 2) have been characterized, and their effects on the skin in Vitro and in ViVo have been evaluated.12 Because of their promising use for skin research, the inclusion of ceramides in bicelles was also previously studied. Note that ceramides are the major lipid constituent of lamellar sheets present in the intercellular spaces of the SC. It has been proven that bicelles support the inclusion of 10 mol % ceramides without affecting their structural integrity.13,14 In this work, we especially addressed the application of a bicellar system which is more suitable for skin penetration at physiological conditions. (5) Kartono, F.; Maibach, H. I. Contact Dermatitis 2006, 54, 303–312. (6) Coderch, L.; López, O.; de la Maza, A.; Parra, J. L. Am. J. Dermatol. 2002, 4, 107–129. (7) López, O.; Co´cera, M.; Walther, P.; Wehrli, E.; Coderch, L.; Parra, J. L.; de la Maza, A. Micron 2001, 32, 201–205. (8) López, O.; Co´cera, M.; Wertz, P. W.; López-Iglesias, C.; de la Maza, A. Biochim. Biophys. Acta 2007, 1768, 521–529. (9) López, O.; Co´cera, M.; Campos, L.; de la Maza, A.; Coderch, L.; Parra, J. L. Colloids Surf., A 2000a, 162, 123–130. (10) López, O.; Walther, P.; Co´cera, M.; de la Maza, A.; Coderch, L.; Parra, J. L. Skin Pharmacol. 2000b, 13, 265–683. (11) López, O.; Co´cera, M.; Coderch, L.; Parra, J. L.; de la Maza, A. Langmuir 2002, 18, 7002–7008. (12) Barbosa-Barros, L.; Barba, C.; Co´cera, M.; Coderch, L.; López-Iglesias, C.; de la Maza, A.; López, O. Int. J. Pharm. 2008, 352, 263–272. (13) Barosa-Barros, L.; de la Maza, A.; Walter, P.; Estelrich, J.; López, O. J. Microsc. 2008, 230 (1), 16-26. (14) Barbosa-Barros, L.; de la Maza, A.; López-Iglesias, C.; López, O. Colloids Surf., A 2008, 317, 576–584.

10.1021/la703732h CCC: $40.75  2008 American Chemical Society Published on Web 05/10/2008

Penetration and Growth of DPPC/DHPC Bicelles

Thus, the effect of bicelles made up of DPPC and DHPC (q ) 3.5 and total lipid concentration CL ) 20%) on the microstructure of the skin was evaluated. This system was characterized by 31P NMR, small-angle X-ray scattering (SAXS), dynamic light scattering (DLS), and freeze fracture electron microscopy (FFEM). FFEM was also used to evaluate in Vitro the microstructural alterations of the SC induced by the DPPC/ DHPC bicelles.

2. Experimental Section 2.1. Materials and Sample Preparation. Bicelles were made up of 1,2-dipalmitoyl-sn-glycero-3-phosphocholine (DPPC) and 1,2dihexanoyl-sn-glycero-3-phosphocholine (DHPC) purchased from Avanti Polar Lipids (Alabaster, AL). DHPC is a short chain phospholipid soluble in water and chloroform and insoluble in acetone. Because of its high hygroscopic feature, this product was acquired in a chloroform solution. The 31P NMR study of bicellar systems required deuterium oxide as the internal standard, which was acquired from Sigma-Aldrich (St. Louis, MO). Trypsin and phosphate buffered solution (PBS), needed for SC isolation, were also purchased from Sigma-Aldrich. Bicellar samples were prepared by mixing appropriate amounts of DPPC powder and DHPC chloroform solution to reach the DPPC/ DHPC molar ratio of q ) 3.5. After mixing the components, the chloroform was removed with a rotary evaporator and the systems were hydrated with water to reach 20% (w/v) of the total lipid concentration. The sample for 31P NMR was hydrated with deuterium oxide instead of water. The solutions were prepared by subjecting the samples to several cycles of sonication and freezing until the samples became transparent. The samples were analyzed by 31P NMR, SAXS, DLS, and FFEM. All the measurements were performed at 37 °C to observe the characteristics of the aggregates at the physiological temperature. Additionally, 31P NMR was also performed at 45 °C in order to observe the ability of the system to align in the magnetic field at temperatures above the Tm of DPPC, which is 41 °C.15 Sections of pieces of native pig SC were placed in water at 65 °C for 4-5 min, and the epidermis was scraped off in sheets. These sheets were placed in 100 mL of 0.5% trypsin in PBS (pH 7.4, 4 °C, overnight). The SC pieces were then collected, rinsed with distilled water, and suspended in a large volume of water. The pieces were transferred to a round flask with fresh trypsin/PBS solution, and the flask was rotated at 100 rpm (25 °C, 2 h) and washed with distilled water.6 Afterward, the SC pieces were incubated with bicelles (treated SC) and deionized water (native SC) for 18 h at 37 °C. The pieces were then washed with distilled water for 30 min and observed via cryogenic scanning electron microscopy (cryo-SEM) following the procedure described in section 2.2.4. 2.2. Experimental Methods. 2.2.1. 31P Nuclear Magnetic Resonance. NMR spectra were recorded on a Varian System 400 MHz spectrometer (Varian NMR Systems, Palo Alto, CA) equipped with a 5 mm probe at room temperature. 31P NMR spectra were recorded at 161.901 MHz using a single pulse, quadrature detection, complete phase cycling of the pulses, and proton decoupling during the signal acquisition. The acquisition parameters were as follows: pulse length of 8.5 µs, recycle delay of 0.5 s, spectral width of 6868.1 Hz, and 22 K data size. Spectra were referenced to 85% phosphoric acid. 2.2.2. Small-Angle X-ray Scattering. SAXS measurements were carried out using a small-angle X-ray Kratky camera (Hecus GmbH, Graz, Austria) coupled to a Siemens KF 760 (3 kW) generator (Siemens AG, Karlsruhe, Germany). Nickel-filtered radiation with a wavelength corresponding with the Cu KR line (1.542 Å) was used. The linear detector was PSD-OED 50 (M-Braun, Garching, Germany), and the temperature controller was a Peltier KPR AP Paar model (Anton Paar, Graz, Austria). The sample was inserted between two Mylar sheets with a 1 mm separation. The SAXS (15) Leonenko, Z. V.; Finot, E.; Ma, H.; Dahms, T. E. S.; Cramb, D. T. Biophys. J. 2004, 86, 3783–3793.

Langmuir, Vol. 24, No. 11, 2008 5701 scattering curves were smoothed with a procedure that includes a third degree polynomial fitting routine.16 This routine was set up to ensure no changes in the slopes or in the peak position and sharpness. The system uses a line collimated beam; therefore, to preserve sharpness, the smoothed curves were desmeared using the procedure of Singh et al.17 The SAXS curves are shown as a function of the scattering vector

Q ) 4π ⁄ λ sin θ

(1)

where 2θ is the scattering angle and λ is the wavelength of the radiation (1.542 Å). The scattering intensity of uncorrelated lipid bilayers is given by

I(q) ) |F(q)|2 ⁄ q2

(2)

where the form factor is the Fourier transform of a simplified Gaussian bilayer model18 and takes the form

F(q) ) (2π)1⁄2[2σH exp(-σH2q2 ⁄ 2) cos(qzH) σCFR exp(σC2q2 ⁄ 2)] (3) where zH and σH are the center and the width, respectively, of the headgroup Gaussian and σC and FR are the width and the relative amplitude, respectively, of the hydrocarbon chain Gaussian. The bilayer thickness dB was then determined directly from the model fit

dB ) 2(zH + 2σH)

(4)

For further details, see the recent review of Pabst.19 2.2.3. Dynamic Light Scattering. The hydrodynamic diameter (HD) and polydispersity index (PI) of the bicellar systems were determined by means of DLS using a Zetasizer Nano ZS instrument (Malvern Systems, Southborough, MA). DLS measures the Brownian motion of the particles and correlates this to the particle sizes. The relationship between the size of a particle and its speed due to Brownian motion is defined in the Stokes-Einstein equation:

HD ) kT ⁄ 3πηD

(5)

where HD is the hydrodynamic diameter, D is the translational diffusion coefficient (m2/s), k is the Boltzmann constant (1.3806503 × 10-23 m2 kg s-2 k-1), T is the absolute temperature (K), and η is the viscosity (mPa s). The particle sizes were determined by detection and analysis of the scattered light from the 632 nm He/Ne laser beam. The noninvasive backscatter technology was used in order to minimize multiple scattering effects. The detection of the light scattered was performed at an angle of 173°. 2.2.4. Freeze Fracture Electron Microscopy. The high-pressure freezing (HPF) technique was carried out as described by Walther and Ziegler.20 A 400 square mesh 3 mm gold grid containing bicellar sample was placed on aluminum planchett sandwiches and highpressure frozen at 2100 bar and -150 °C using a Compact HPF-01 instrument (Wohlwend Engineering, Sennwald, Switzerland). All objects and equipment used in the preparation procedure were acclimated at 37 °C in order to freeze the samples from this temperature. After freezing, the sample was removed from the holder and stored under liquid nitrogen.21 The frozen aluminum planchett sandwiches were mounted on a Bal-Tec holder and transferred to the BAF 300 freeze-etching device for freeze fracture. The fracture was produced at -150 °C and under vacuum of about 2 × 10-7 (16) Plaza, M.; Tadros, Th. F.; Solans, C.; Pons, R. Langmuir 2002, 18, 5673– 5680. (17) Singh, M.; Ghosh, S.; Shannom, R. J. Appl. Crystallogr. 1993, 26, 787– 794. (18) Pabst, G.; Koschuch, R.; Pozo-Navas, B.; Rappolt, M.; Lohner, K.; Laggner, P. J. Appl. Crystallogr. 2003, 36, 1378–1388. (19) Pabst, G. Biophys. ReV. Lett. 2006, 1, 57–84. (20) Walther, P.; Ziegler, A. J. Microsc. 2002, 208, 3–10. (21) Walther, P.; Wehrli, E.; Hermann, R.; Mu¨ller, M. J. Microsc. 1995, 179, 229–237.

5702 Langmuir, Vol. 24, No. 11, 2008 mbar. The replicas were obtained by unidirectional shadowing with 2 nm of Pt/C at an angle of 45° and a backing layer of 20 nm of C perpendicularly. In the cases of native SC and SC treated with bicelles, small pieces of SC (total thickness of about 150 µm) were immersed in 1-hexadecene, placed on aluminum planchettes using an ophthalmologic punch, and covered with other aluminum planchett forming sandwiches. n-Hexadecene is a hydrophobic and chemically inert paraffin oil with low viscosity and surface tension, which improves the transfer of both pressure and cold to the specimen by occupying the empty space between the sample and the planchett. The samples were then high-pressure frozen in a similar way as with the bicelles. The aluminum planchett sandwiches were transferred to the BAF 300 freeze-etching device. The temperature of the sample stage was raised to about 120 °C (vacuum at about 2 × 10-7 mbar). Samples were cryo-sectioned using a special diamond knife (manufactured by the GFD, Gesellschaft fu¨r Diamantprodukte, Ulm, Germany).22 After sectioning, the samples were kept in the vacuum for 60 s in order to allow some water to sublimate (“etching”). The samples were then double-layer coated. Platinum-carbon (3 nm) was electron beam evaporated at an angle of 45°.21 A backing layer of 5 nm of carbon was evaporated perpendicularly. In all cases, after coating, the holder was removed, mounted onto the Gatan cryo-holder 626 (Gatan, Inc., Pleasanton, CA), and transferred to the cryogenic scanning electron microscope. Specimens were investigated at a temperature of -115 °C in the Hitachi S-5200 in-lens field-emission scanning electron microscope (Hitachi, Tokyo, Japan). Imaging was performed by collecting the backscattered electron (BSE) signal. BSEs were directly recorded with the built-in YAG-BSE detector and (indirectly) with the converted BSE signal, the so-called “composite rich image”. To improve the signal tonoise ratio, the YAG-BSE image and the composite rich image were superimposed. The digital images were processed with Adobe Photoshop. No other image processing apart from brightness and contrast correction was performed.

Barbosa-Barros et al.

Figure 1. 31P NMR spectra of DPPC/DHPC bicelles measured at 37 and 45 °C. The spectra were recorded at 161.901 MHz using a single pulse, quadrature detection, complete phase cycling of the pulses, and proton decoupling during the signal acquisition.

3. Results 3.1. Bicelle Alignment Studied by 31P NMR. The 31P NMR technique is commonly used to study structural and conformational aspects of the bicellar systems. 31P NMR resonance line shapes and shifts contain information concerning the headgroup conformation, the membrane morphology, and the orientation of the lipid bilayer with respect to the magnetic field.23 This technique allows for the characterization of disklike bicelles and their ability to align in a magnetic field, which requires a transition from the gel phase to the fluid liquid-crystalline phase (above Tm). The aligned samples provide an anisotropic environment that produces a characteristic spectrum composed of two well-differentiated resonances, whereas the nonaligned samples present a low field resonance peak or doublet.24 31P NMR spectra were obtained at 37 °C (below the Tm of DPPC) and at 45 °C (above the Tm of DPPC) (Figure 1). The reference peak is shown at 0 ppm. At 37 °C, the spectrum reveals a doublet near -0.6 ppm with no orientation characteristics (Figure 1, bottom line). These closed spaced peaks reflect the slightly different environments of the DPPC and DHPC lipids in the bicelle structure.25 At 45 °C, the two differentiated resonances at 0.6 and -1.6 ppm indicate the magnetic alignment of the sample (Figure 1, top line). 3.2. Bicelle Dimensions and Morphology: SAXS, DLS, and FFEM. The small-angle X-ray diffraction pattern of the bicelles is plotted in Figure 2. The continuous line gives the best fit for the curve, applying the form factor squared of the simplified Gaussian model (eqs 2 and 3) described in section 2.2.2. This model accounts for the general features of the electron density (22) Walther, P. Microsc. Microanal. 2003, 9, 279–285. (23) Hemminga, M. A.; Cullis, P. R. J. Magn. Reson. 1982, 47, 307–323. (24) Marcotte, I.; Auger, M. Concepts Magn. Reson. 2005, 24, 17–37. (25) Triba, M. N.; Warschawski, D. E.; Deveaux, P. F. Biophys. J. 2005, 88, 1887–1901.

Figure 2. Scattered intensity of the DPPC/DHPC bicelles measured at 37 °C. The continuous line represents the best fit of the squared form factor model (eqs 2 and 3). For details, see the Experimental Methods section.

modulation along the z-direction and requires the adjustment of only four parameters.26 The fit results were as follows: zH ) 2.1 nm, σH ) 0.32 nm, FR ) 0.87, and σC ) 0.41 nm. According to eq 4, a value of bilayer thickness dB ) 5.4 nm is obtained from the electron density profile. The size distribution curve obtained by DLS (Figure 3) shows a hydrodynamic diameter (HD) of 11.3 nm and a polydispersity index (PI) of 0.072 for this sample. Through this technique, the HD of a hypothetical hard sphere that diffuses with the same speed as the particle under experiment is given. As the bicellar structure is disk-shaped, the particle size obtained by DLS provides a relative measurement of the structure dimensions. Therefore, this data should be interpreted taking into consideration also the FFEM images, in which a direct (26) Rappolt, M.; Laggner, P.; Pabst, G. Recent Res. DeV. Biophys. 2004, 3, 363–392.

Penetration and Growth of DPPC/DHPC Bicelles

Langmuir, Vol. 24, No. 11, 2008 5703

Figure 3. DLS size distribution curve of the DPPC/DHPC bicelles. CL ) 20% as measured by detection and analysis of scattered light at an angle of 173°. The average particle size at 37 °C corresponds to a HD of 11.3 nm and PI ) 0.072.

Figure 4. Cryo-SEM image of the DPPC/DHPC bicelles at 37 °C. The black arrows denote the disks viewed face-on, and the white arrows indicate disks viewed edges-on.

visualization of the bicellar structures is obtained. Figure 4 shows small discoidal aggregates of about 15 nm in diameter. In this image, the face (black arrows) and the edges (white arrows) of the disks are visualized. Hence, this result shows quite good agreement with the DLS data despite the different resolution of each technique. For well-dispersed samples, the accuracy and precision of a measurement with a Zetasizer Nano ZS instrument is within 2%.27 The main advantage of DLS over FFEM is that the particle size is obtained in minutes. However, misleading results may be obtained in the averaging process. FFEM measurements have the appeal that one obtains directly an image of the feature whose size is to be determined.28 Nevertheless, it is important to consider that the coating process involves the deposition of a 2 nm layer of platinum on the sample. As a consequence, an associated error of 0-2 nm in the size measurement of the objects should be considered. In order to check the accordance of the results obtained by these two techniques, it is possible to combine the diameter as obtained from FFEM and the bilayer thickness (obtained by SAXS) to calculate the hydrodynamic diameter of the particles. Taking the diameter to be 15 nm and the thickness as 5.4 nm, we can approximate the form to an oblate ellipsoid with a major axis of 15 nm and a minor axis of 5.4 nm. According to the literature,29 HD ∼ 0.95 DS, where DS is the diameter of (27) Malvern technical note MRK728-01, Malvern Instruments, http:// www.malvern.co.uk/common/downloads/campaign/MRK728-01.pdf. (28) Severs, N. J.; Shotton, D. M. Rapid freezing, freeze fracture, and deep etching; Wiley-Liss, Inc.: New York, 1995; Chapter 12. (29) Schmitz, K. S. An Introduction to Dynamic Light Scattering by Macromolecules; Academic Press: San Diego, CA, 1990; p 50.

Figure 5. Micrographs obtained by cryo-SEM of the native stratum corneum replicas of samples cryofixed using high-pressure freezing. The fracture plane lies perpendicular to the skin surface. In part (A), the white arrow points to the corneocyte area and the black arrow points to the intercellular lipid area evidencing a fracture across the lamellar structures. Part (B) shows the corneocyte cross-fractured area evidencing the granular pattern of keratin.

the sphere with equivalent volume to the disk. With the present dimensions, this calculation results in HD ) 11.6 nm, in excellent agreement with the light scattering results. 3.3. Structure of Native SC and Bicelle Treated SC. The SC samples were high-pressure frozen from 37 to -150 °C and visualized by cryo-SEM. The great advantage of HPF is the nonrequirement of any sample pretreatment such as chemical fixation and staining, which frequently leads to the formation of artifacts. Moreover, no replica cleaning is requested for cryoSEM (reviewed by Echlin30). Replica cleaning for transmission electron microscopy (TEM) is tedious and sometimes difficult; large replicas tend to disintegrate during cleaning, and the field of view is restricted by the grid bars. In SEM, the sample does not have to be transparent to the electron beam, and bulk samples can be analyzed. Using the combination of HPF and cryo-SEM, the bulk freeze fractured samples were imaged and large areas could be investigated. The micrographs obtained for the native SC are shown in Figure 5. In these images, the fracture plane lies perpendicular to the skin surface. Panel A displays the corneocyte area (white arrow) and the lipid intercellular spaces evidencing a fracture across the lamellar structures (black arrow). Panel B shows a cross fracture of the corneocytes characterized by the particular granular pattern of keratins entirely filling their (30) Goldstein, J. I.; Newbury, D. E.; Echlin, P. Scanning electron microscopy and X-ray microanalysis, 2nd ed.; Plenum Press: New York, 1992.

5704 Langmuir, Vol. 24, No. 11, 2008

Barbosa-Barros et al.

Figure 7. Evolution of the average particle size of DPPC/DHPC bicelles with q ) 3.5 by the effect of dilution as measured by DLS. The polydispersity indices (PIs) are shown in the inset.

Figure 6. Micrographs obtained by cryo-SEM of the treated stratum corneum replicas of samples cryofixed using high-pressure freezing. In part (A), the white arrow shows vesicle structures with sizes around 200 nm and the arrowhead points to the lamellar-like structures in the intercellular lipid areas. The black arrow shows a corneocytes area. Part (B) displays the magnification of vesicles in the intercellular lipid area (white arrows) between two cross-fractured corneocytes (black arrows).

interior and by the absence of cell organelles. The micrographs of SC treated with bicelles are shown in Figure 6. No changes were detected in protein domains (black arrows) when compared with the native SC. Nevertheless, some structural changes were found in the lipid lamellae regions. The image in 6A shows vesicle structures with sizes around 200 nm (white arrow) and lamellar-like structures with irregular shapes (black arrowhead) in the lipid zones of the SC. Figure 6B displays a high magnification image of vesicles in the lipid lamellae areas. In this image, the vesicles (white arrows) are visualized between two cross-fractured corneocytes (black arrows). These structures were not detected in the native SC, indicating that their appearance is related to the treatment of the SC with the DPPC/DHPC bicelles. 3.4. Bicelle-Vesicle Transition by Dilution. To monitor the effect of dilution on the structure of the bicelles, 2 mL of the DPPC/DHPC (q ) 3.5, CL ) 20%) sample was sequentially diluted with deionized water in seven steps (D1-D7) to the following concentrations: D1 ) 5%, D2 ) 2.5%, D3 ) 1.25%, D4 ) 0.62%, D5 ) 0.31%, D6 ) 0.15%, and D7 ) 0.07%. Twenty-four hours after the dilutions, the hydrodynamic diameter (HD) and polydispersity index (PI) of each diluted sample were measured by DLS at 37 °C. An overview of this process is shown in Figure 7. The DLS curves of the diluted samples show that the HDs of the structures increase upon dilution from small values

of 11.3 nm assigned to bicelle disks to large size aggregates bigger than 1 µm, indicating structural transitions in the sample. For better understanding of the results, aside from the evolution of the average particle sizes, a detailed analysis of the scattering intensity was performed. The particle sizes obtained were separated in three groups: P1 includes particle sizes in the range 10-100nm, P2 includes intermediate particle sizes ranging from 101 to 500 nm, and P3 includes high values of particle sizes >500 nm. The HD and the percentage of the light scattered of each group were plotted separately as a function of the sequential sample dilution (Figure 8). From this figure, we observe that the intensity for small bicelles (P1) diminishes with dilution whereas the HD increases (Figure 8A). Intermediate aggregates (P2 in Figure 8B) are present from the first dilution on and display a maximum of intensity and size at D3. From D4, the intensity of these aggregates diminishes and bigger structures are detected (P3 in Figure 8C), coexisting with P1 and P2 until D6. The high PI values (shown in Figure 7) from D3 onward are explained by the high variety of aggregates present in these samples. It is noteworthy that the percent of the intensity curves does not represent the percentage of the structures present in the systems. In DLS, the HD is calculated from the intensity of the scattered light and gives information about the different particles present in the sample. From the Rayleigh approximation follows that the intensity is proportional to d6 (d is the particle diameter); thus, the contribution of the light scattered from small particles is relatively small when compared to that of large particles that scatter much more light. Hence, the intensity curves obtained indicate the appearance of bigger aggregates by dilution but do not accurately quantify them. In general, the percent intensity of these structures and their HD increase in each dilution (Figure 8). The coexistence of these big aggregates with the small bicelles is detected until the last dilution performed (D7). In order to investigate the morphology of these systems, the diluted bicellar solutions were analyzed by FFEM. A representative micrograph of the sample D5 is shown in Figure 9. This image reveals the presence of vesicles of about 200-500 nm (black arrows) together with small bicelles (white arrows). Comparing this micrograph to that of the original system (Figure 4), the increase of structure sizes and the variety of bigger aggregates in the sample are noteworthy. This result is fully consistent with DLS data corroborating the transition of bicelles from disks to vesicles by the effect of dilution. This transition took place by progressive steps that implied the coexistence of different aggregate structures in the medium.

Penetration and Growth of DPPC/DHPC Bicelles

Langmuir, Vol. 24, No. 11, 2008 5705

Figure 9. Cryo-SEM image of the diluted DPPC/DHPC bicelles (D5) at 37 °C. The black arrows denote vesicle structures of about 200-500 nm, and the white arrows point to bicellar structures.

Figure 8. Evolution of the DLS intensity and hydrodynamic diameter of the DPPC/DHPC bicelles upon dilution. P1 denotes the evolution of particles with sizes in the range of 10-100 nm (A), P2 denotes intermediate sized particles with diameters around 100-500 nm (B), and P3 indicates the bigger particles of >500 nm (C).

4. Discussion Typical binary mixtures of bicelles are composed of DMPC as the long chain lipid and DHPC as the short chain lipid. Although this classical model has been successfully used in several studies, bicelles made up of lipids with different acyl chains, backbones, or headgroups can offer alternatives for a number of studies.31–35 DPPC is one of the most studied lipids for bilayer models. Because of its longer acyl chains, the bilayer thickness of DPPC provides a better model for biological membranes.36 For our purposes, DPPC has the additional advantage of having a higher value of Tm (41 °C) when compared with DMPC (23 °C). In bicellar aggregates, phase transitions take place from Tm.37–39 These transitions involve morphological changes in the bicellar structures, that is, from disks to cylindrical micelles to perforated lamellar sheets and mixed multilamellar vesicles.40 In order to obtain a system able to penetrate through the narrow intercellular (31) Ottiger, M.; Bax, A. J. Biomol. NMR 1999, 13, 187–191. (32) Cavagnero, S.; Dyson, H. J.; Wright, P. E. J. Biomol. NMR 1999, 13, 387–391. (33) Struppe, J.; Whiles, J. A.; Vold, R. R. Biophys. J. 2000, 78, 281–289. (34) Whiles, J. A.; Glover, K. J.; Vold, R. R.; Komives, E. A. J. Magn. Reson. 2002, 158, 149–156. (35) Aussenac, F.; Lavigne, B.; Dufourc, E. J. Langmuir 2005, 21, 7129–7135. (36) Tiburu, E. K.; Moton, D. M.; Lorigan, G. A. Biochim. Biophys. Acta 2001, 1512, 206–214. (37) Lewis, R. N. A. H.; Mak, N.; Edhaney, R. N. M. Biochemistry 1987, 26, 6118–6126. (38) Laughlin, R. G.; Munyon, R. L.; Fu, Y.-C.; Fehl, A. J. J. Phys. Chem. 1990, 94, 2546–2552. (39) Marsh, D. CRC Handbook of Lipid Bilayers; CRC Press: Boca Raton, FL, 1990.

spaces of the skin SC, bicelles of small size at physiological temperature are needed. This requirement is fulfilled by DPPC/ DHPC q ) 3.5 bicelles. At 37 °C, below the DPPC Tm, these structures have dimensions of about 15 nm in diameter and 5.4 nm in thickness. These values are consistent with previously reported data of DPPC bilayer thickness41 and bicellar disk dimensions.42 In addition, 31P NMR spectra of this system accurately demonstrate its ability to align in the magnetic field at temperatures above Tm, corroborating the existence of disk-shaped structures in the sample. The passage of these small bicelles through the SC lipid region seems reasonable, considering that this region is formed by lipid lamellae with narrow interlamellar spaces (between 6 and 10 nm).6 In our FFEM images of the treated SC, it was not possible to track the presence of the bicellar disks, but, interestingly, vesicles and lamellar-like structures were observed in the lipid intercellular areas. These structures were probably the result of a structural rearrangement of bicelles inside the cutaneous tissue. DLS and FFEM analyses of the bicellar diluted samples demonstrate the tendency of the bicellar aggregates to grow and form vesicles by the effect of dilution (Figures 7 and 8). It is noteworthy that the phase transitions the bicelles underwent due to the variation of lipid concentration, temperature, and phospholipid molar ratio43 are very similar to those involved in the reconstitution of the surfactant-lipid micellar systems.44 The transformation of these structures in vesicles was largely discussed in a number of works.45 In our earlier studies, we investigated some kinetics aspects of this process.46 A model for the micelleto-vesicle transition proposed by Leng et al. describes the rapid formation of disklike aggregates and their growth and closure to form vesicles.47 This model takes into account line tension dominating bending energy. Certainly, the resemblance of surfactant-lipid micelles and phospholipid bicelles justifies their similar behavior. DHPC solubilizes the DPPC bilayer forming the bicellar structures in a similar way that a surfactant solubilizes lipid vesicles forming micelles. In addition, the high water (40) Harroun, T. A.; Koslowsky, M.; Nieh, M.-P.; de Lannoy, C.-F.; Raghunathan, V. A.; Katsaras, J. Langmuir 2005, 21, 5356–5361. (41) Nagle, J. F.; Tristam-Nagle, S. Biochim. Biophys. Acta 2000, 1469, 159– 195. (42) Ottiger, M.; Bax, A. J. Biomol. NMR 1998, 12, 361–372. (43) Barbosa-Barros, L.; de la Maza, A.; Estelrich, J.; Linares, A. M.; Feliz, M.; López, O. Biophys. J. 2007, 92(4), 234a. (44) Pinaki, R. J. Phys. Chem. B 2002, 106, 10753–10763. (45) Goltsov, A. N.; Barsukov, L. I. J. Biol. Phys. 2000, 26, 27–41. (46) López, O.; Co´cera, M.; Coderch, L.; Parra, J. L.; Barsukov, L.; de la Maza, A. J. Phys. Chem. B 2001, 105, 9879–9886. (47) Leng, J.; Egelhaaf, S. U.; Cates, M. E. Biophys. J. 2003, 85, 1624–1646.

5706 Langmuir, Vol. 24, No. 11, 2008

solubility of DHPC accounts for the structural changes the bicelles underwent. When DHPC is removed from the bicelles upon dilution (q increases), the bilayers tend to fuse and the bicellar diameter increases. Morphological changes in the bicelle structure then take place. When sufficient DHPC is removed from the bicelles, the precipitation of large aggregates is visible, that is, phase separation occurs.48 From these findings, we attribute the presence of vesicles in the intercellular spaces of the SC to a structural transition process from bicelles to vesicles occurring inside of this tissue. The solution used to treat the SC was composed of bicelles with dimensions suitable for SC penetration. Inside this tissue, bicelles probably transformed in vesicles following a process analogous to that observed in the DLS experiment with diluted bicellar solutions. A similar phenomenon was previously reported by our group applying octylglucoside/phosphatidylcholine (OG/ PC) mixed micelles in the SC.11 In this earlier investigation, we demonstrated that liposomes were solubilized by OG and the resulting mixed micelles penetrated through the SC and then were reconstituted in vesicles by effect of the hydration gradient of the skin. Taking into account that in the present work the SC pieces were washed with distilled water for 30 min after incubation with bicelles, a dilution of the bicelles inside the SC has to be considered in order to explain the vesicle formation. The penetration of water in the SC pieces probably promoted this dilution, giving rise to the appearance of the vesicles. Another possibility is that the interaction of bicelles with the complex hydrated lipid environment of the SC layers would have promoted the morphological transitions observed in the aggregates. The presence of the lamellar-like structures in the SC intercellular areas could be the result of the interaction between the lipids from bicelles and the lipids of the SC intercellular spaces. Although we have no experimental evidence to explain how these structures formed, their presence requires a previous diffusion of the lipids from the bicelles. The vesicles probably interacted with the lipid areas of the intercellular spaces and formed a lamellar structure similar to that present in the native SC, which can be observed in the cryo-SEM images. In summary, bicelles present several advantages as a possible vehicle for skin penetration when compared with lipid-surfactant systems: (1) the bicelle structure contains a bilayer that allows for the incorporation of substances; (2) the presence of DHPC controlling the bicelle diameters allows modulation of aggregate sizes; and (3) bicelles are composed of lipids only. The use of surfactants is unadvisable because these molecules damage the skin barrier function, provoking the breaking of the corneocyte envelopes and disorganizing the intercellular lipid structures.5 Our micrographs of both native SC and treated SC show preserved cellular and intercellular structures. This indicates that bicelles interact with SC structures without promoting microstructural damage. Moreover, the lamellar-like structures shown in our micrographs account for a reinforcement effect of the lipid intercellular spaces by the treatment with bicelles. This effect may be useful for possible applications on the preservation and/ or regeneration of the barrier function of the skin. Recently, we also studied the interaction of DMPC/DHPC bicelles with a molar ratio q ) 2 in the skin in Vitro and in ViVo.12 In that work, we demonstrated that bicelles increase permeability and elastic parameters without promoting irritation of the skin in ViVo. Nevertheless, the in Vitro analysis of the SC incubated with those bicelles did not show vesicle formation or lamellaelike structures as we detected in the present work. This result can be explained by the lower Tm of DMPC (23 °C). The temperature (48) Struppe, J.; Vold, R. R. J. Magn. Reson. 1998, 135, 541–546.

Barbosa-Barros et al.

at which the SC was incubated with the DMPC/DHPC bicelles was the same (37 °C). At this temperature, the system was in the fluid phase, therefore forming large aggregates.13 These large aggregates could not penetrate through the narrow intercellular spaces of the SC, and their effects were concentrated in the skin surface. Indeed, no microstructural change in the SC was visualized in that investigation. By contrast, the bicelles made up of DPPC/DHPC used in the present investigation demonstrated their suitability for skin penetration. As the Tm of DPPC is 41 °C, the system was in the gel phase and was composed of small aggregates at the experimental temperature of incubation with the SC (37 °C). In addition, the higher long/short chain molar ratio of these bicelles (q ) 3.5) represents an advantage when compared with the DMPC/DHPC bicelles of q ) 2 due to the increased presence of bilayer-forming lipids in the medium. The chance to modulate the bicellar structure is one of the principal advantages of these nanoaggregates. This fact makes possible the use of bicelles for diverse purposes. The penetration and growth of bicelles inside the SC opens up new avenues in the application of these systems. Bicelles accurately fulfill the requirements for an effective skin carrier due to their size, structure, and composition. Although these aggregates have no aqueous internal compartment for encapsulating drugs, their bilayered structure allows for the inclusion of lipophilic and amphiphilic compounds in their structures. Moreover, because of the bicelles’ capacity to increase the permeability parameters of the SC,12theycouldalsoactasanenhancerforthepenetrationofhydrophilic components dissolved in the aqueous medium. Further, the conversion of bicelles into vesicles inside the SC hinders their migration outside the tissue and allows for a lipid reinforcement effect in the skin. In addition, this property could be very useful to intensify the effect of specific compounds carried by bicelles in the SC layers. In ViVo, the driving force for the rearrangement of the bicellar lipids would be the hydration gradient across the skin, which varies from 15 to 29% in the SC and reaches 70% in the stratum granulosum. A profound examination of the interaction of these bicelles and the skin in ViVo will be necessary to obtain a complete view of the potential of these structures for skin purposes. This is the subject of our current investigations.

5. Conclusions The DPPC/DHPC bicelles studied were able to penetrate the skin SC in Vitro and to grow, forming vesicles inside the intercellular lipid spaces. This growth was also observed in the dynamic light scattering experiments when bicelles were diluted with water. The morphological changes these structures underwent resemble micelle-to-vesicle transitions of the lipid-surfactant systems. Because of the absence of surfactant in their composition, and because of their size and slim shape, bicelles present great advantages to other lipid systems: they show good penetration ability and high skin compatibility. Moreover, the possibility of adapting the aggregates’ morphology depending on the specific application makes the bicelles a smart nanosystem. These aggregates could represent a new skin-compatible carrier for drug delivery, enriching the skin per se. Their conversion in vesicles allows for the retention of substances inside the SC intercellular spaces. In this way, future works must be directed to the optimization of these systems, especially regarding their stability and drug entrapment capacity. Acknowledgment. The authors wish to thank Jaume Caelles, Carmen López-Iglesias, and Eberhard Schmid for technical support. This study was financed in part by a research grant subsidized by the German Academic Exchange Service (DAAD). LA703732H